ABSTRACT The mechanisms involved in brain drug interactions and reactive metabolites distribution in the brain during pharmacotherapy is largely unknown. Cytochrome P450 (CYP) drug-metabolizing enzymes are known to be functionally active at the blood-brain barrier (BBB). Multidrug transporter proteins (e.g., MDR1) at the BBB were once thought to be the main mechanism of failed antiepileptic drug (AED) therapy in patient suffering from drug-resistant epilepsy (DRE). However, we reported that the BBB participates in this process via CYP-mediated mechanism through the conversion of a therapeutic agent into a neurotoxic molecule. Increased CYP-MDR1 co-expression was found in epileptic brain endothelial cells (EPI-ECs) and neurons. We also found neurotoxic interactions between carbamazepine, (CBZ, an AED) and sertraline (antidepressant) ? a consequence of CYP-mediated production of reactive CBZ metabolites and oxidative stress. The extent, relevance and topography of co-prescribed drug metabolism and potential neurotoxicity in the epileptic brain remain unexplored. Some of these drugs are induced and/or inhibited by CYP; we must devise ways to prevent subsequent neuronal changes. Our multidisciplinary team will test the central hypothesis that two co-prescribed AEDs will competitively interact in human BBB endothelial cells to regulate neuronal function and a CYP-MDR mechanism affects drug- metabolite levels in DRE brain. We will study a homogeneous patient population (with focal DRE due to cortical dysplasia, FCD) co-prescribed combination of CYP- and/or non-CYP-regulated AEDs. The Specific Aims will be achieved by using these multifaceted approaches to: (1) assess CYP?s contribution to neurotoxicity and metabolism of AEDs in a humanized in vitro neurovascular unit (NVU); (2) determine the regulatory factors and the extent to which CYP metabolize AEDs by using (a) EPI-ECs derived from brain resections of DRE with FCD to form a NVU and (b) epileptic (dysplastic) and nonepileptic (nondysplastic) tissue resected from the individual patient (ex vivo); and (3) study in situ CYP/MDR1 expression correlated to AED tissue bioavailability, formation of reactive metabolites, and extent of cellular damage in epileptic vs. nonepileptic resected DRE brain regions. The human tissue will be characterized histologically and electrophysiologically through invasive SEEG or subdural grid recording. A novel humanized NVU with primary control or patient-derived DRE brain cells will be used. Drug interactions at the BBB and real-time neuronal effects will be evaluated as a ?personalized medicine? tool for disease modeling. Further measurement includes drug/metabolites by HPLC-MS (pharmacokinetics); neurotoxicity; free radical assay; and mitochondrial (pharmacodynamics) and gene expression changes (pharmacogenomics). Molecular manipulations of brain cells (by transfection/inhibitors) will validate CYP-mediated AED interactions in normal, drug-resistant BBB and CYP-MDR1 regulation in DRE brain (epileptic/nonepileptic). Together, there will be ample information to minimize CYP-dependent neurotoxicity and prevent side-effects of a polytherapy.